(1). Effect of cytokine gene transduction on the tumorigenicity and immunogenicity of tumor cells: In testing two different murine tumor lines, we found that transduction of B16 melanoma with GM-CSF gene rendered the cells to be more immunogenic. In contrast, transduction of NG fibrosarcoma with GM-CSF gene showed increased tumorigenicity but did not change the immunogenicity. The increase of tumorigenicity was attributed both to the increased inherent porliferative ability of the tumor cells and the in vivo production of tumor enhancing growth factor(s) at tumor site, probably through the interaction of GM-CSF with macrophages. (2). Adoptive immunotherapy with anti-CD3-induced activated killer cells of B cell lymphoma: The anti-CD3-induced activated killer cells, CD3-AK, were predominantly CD8(+) cells which selectively killed tumor targets. We found that in the presence of anti-CD3 antibody, the killing of B lymphoma cells was significantly augmented. This was due to the fact that the Fab portion of anti-CD3 will bind to the CD3 molecules of the CD3-AK cells which are cross-linked to the B lymphoma cells that express the Fc receptors to bind to the Fc portion of anti- CD3 and thus increase the efficiency of targeting. Furthermore, engagement of CD3 molecules with the antibody serves as a strong triggering signal to increase the cytolytic activity of CD3-AK cells. When the CD3-AK cells were used as the anti-tumor effectors in adoptive immunotherapy of B cell lymphoma/leukemia, it was found that 25% long- term survival was obtained. With a more vigorous treatment protocol, a 80% long term cure rate was achieved. In view of the fact that there is no need to prepare idiotypic (Id) protein or anti-Id, the use of CD3-AK may be an ideal alternative choice for immunotherapy of B cell lymphoma. (3). Tumor cell-induced immunosuppression: Cancer immunotherapy in clinical trials has been implemented for almost two decades and has met with limited success. The recent attempt of combining the treatment with gene therapy is promising but is still far from being successful. The problem may not be the efficiency of the treatment per se, but may be in the tumor cells which can establish an elaborate escape mechanism to evade host defense system. It is known that many tumor cells lack the recognizable antigenic determinants, MHC, or B-7 which are essential for inducing a host T cell response. The Tab protein may also be defective. Furthermore, tumor cells can induce immunosuppression by actively producing suppressor factors or suppressor cells, or induce T cell anergy/clonal deletion. In studying the effusion associated lymphocytes (EAL) from lung cancer patients with malignant pleural effusion, we found that CTL anergy was induced in vivo in these patients at localized tumor site. The cytolytic activity of the anergized CTL could be restored by activation of the EAL with both IL-2 and anti-CD3 antibody but not by either one alone. The restored cytolytic activity was specifically against the autologous tumor cells and was abrogated by anti-MHC class I antibody. These findings indicate that tumor cells can indeed induce T cell anergy in cancer patients. However, proper manipulation of the anergized T cells can fully restore their function and thus may be useful for adoptive immunotherapy of this tumor.